README.md

metadata

title: Book Recommendation System
emoji: 📚
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 3.37.0
app_file: app.py
pinned: false

📚 Content-Based Book Recommendation System 📖

This is a content-based book recommendation system that recommends books 📕📗 similar to an input book title based on the similarity of book summaries. The system uses TF-IDF (📊 Term Frequency-Inverse Document Frequency) and Cosine Similarity 🧮 to compare books and find the most relevant recommendations. It provides a user-friendly interface built with Gradio 💻, where users can enter a book title and get recommendations.

📂 Project Structure

.
├── app.py                         # 🚀 Main script that runs the app
├── utils.py                       # 🛠️ Helper functions (data loading, model loading)
├── data/
│   ├── books_summary.csv          # 📑 Actual dataset
│   ├── cleaned_books_summary.csv  # 🧹 Preprocessed dataset
├── model/
│   ├── tfidf_vectorizer.pkl       # 🤖 Pre-trained TF-IDF vectorizer
│   ├── tfidf_matrix.pkl           # 🗂️ Pre-calculated TF-IDF matrix
├── src/                           # 📦 Source code folder
│   ├── data_loader.py             # 📥 Module to load and preprocess data
│   ├── feature_engineering.py     # 🧬 Module to create TF-IDF/embedding vectors
│   ├── similarity_calculator.py   # 🧮 Module to calculate similarity matrix
│   ├── recommender.py             # 📚 Main logic to generate recommendations
│   ├── utils.py                   # ⚙️ Utility functions (e.g., cleaning text)
├── requirements.txt               # 📜 List of Python dependencies
└── README.md                      # 📝 Project overview and setup instructions

🌟 Features

• 📚 Book Recommendation: Enter a book title, and the system will recommend the top 5 similar books based on their summaries.
• 🏷️ Categorization: Each recommended book displays its categories as clickable buttons for better user experience.
• 💻 Interactive UI: Simple and clean interface using Gradio.
• 🔧 Modular Code: Functions for data loading, preprocessing, model training, and similarity calculation are separated into different files.

💻 Technologies Used

• 🐍 Python: Core language used to build the system.
• 💻 Gradio: For creating a web-based user interface.
• 📊 Scikit-learn: For TF-IDF Vectorization and Cosine Similarity calculation.
• 🗂️ Pandas: For data manipulation and preprocessing.
• 🔢 NumPy: For numerical operations.

⚙️ Setup Instructions

1. 🧬 Clone the Repository

git clone https://github.com/ajayansaroj17/book_title_recommender.git
cd book_title_recommender

2. 📦 Install Dependencies

Make sure you have Python 3.7+ installed. Then, create a virtual environment and install the required libraries:

python -m venv venv
source venv/bin/activate  # On Windows, use `venv\Scripts\activate`
pip install -r requirements.txt

3. 📚 Download or Prepare the Dataset

The dataset should contain columns for book_name, summaries, and categories. Store the preprocessed dataset as cleaned_books_summary.csv in the data/ folder.

4. 🏋️‍♂️ Pre-train the Model

Run the following script to train the TF-IDF model and create the TF-IDF matrix:

python train_tfidf_model.py

This will:

• 🧠 Train the TF-IDF vectorizer on the summaries column.
• 🗂️ Create the TF-IDF matrix for all books.
• 💾 Save the trained model and matrix as model/tfidf_vectorizer.pkl and model/tfidf_matrix.pkl.

5. 🚀 Run the Application

Launch the Gradio-based web interface by running:

python app.py

The application will open in your browser, allowing you to enter a book title and receive recommendations.

🤔 How the System Works

1. 👤 User Input: The user enters a book title in the input field.
2. 🔍 Recommendation Logic:
   • The system searches for the input book in the dataset.
   • It calculates the TF-IDF vector of the input book's summary and compares it with the summaries of all other books using cosine similarity.
   • The top 5 books with the highest similarity scores are returned.
3. 📊 Output: Recommendations are displayed, including book titles, summaries, and categories as clickable buttons.

📄 File Descriptions

• app.py: 🚀 Main script launching the Gradio UI and handling book recommendations.
• utils.py: 🛠️ Helper functions for loading models, data preprocessing, and utilities.
• feature_engineering.py: 🧬 Trains the TF-IDF model and creates the TF-IDF matrix.
• data/cleaned_books_summary.csv: 📚 Cleaned dataset used for training.
• model/tfidf_vectorizer.pkl and model/tfidf_matrix.pkl: 🤖 Pre-trained TF-IDF model and matrix.

📦 Dependencies

Install the following Python packages using:

pip install -r requirements.txt

• 💻 gradio: For the web interface.
• 📊 sklearn: For TF-IDF and cosine similarity calculations.
• 🗂️ pandas: For data manipulation.
• 🔢 numpy: For numerical operations.

🚀 Potential Extensions and Improvements

• 🏷️ Category-Based Filtering: Filter recommendations by specific categories.
• 🤖 Advanced NLP Techniques: Use embeddings like Word2Vec, GloVe, or transformer-based models like BERT.
• 👥 Personalization: Implement a user profiling system for personalized recommendations.
• ⚡ Scalability: Use Approximate Nearest Neighbors (ANN) for faster similarity calculation on large datasets.

🏁 Conclusion

This project demonstrates building a content-based book recommendation system using TF-IDF and cosine similarity. The modular design ensures easy maintenance and extension, while Gradio simplifies deployment and user interaction!🚀